The proposed application of polymer chain statistics, or configurational statistics, to model carcinogen-nucleic acid covalent complexes is intended to provide an essential first step in relating the subtle features of chemical architecture to the unique physical and biological properties of these systems. Through a combination of semiemprical energy calculations and statistical mechanical analyses, it is now possible to obtain reliable predictions of the secondary and tertiary structures of various nucleic acid systems of established biological significance. A primary objective of this research program then is to elucidate by use of the above techniques the unique conformational features of nucleic acids chemically modified by metabolites of 2-aminofluorene and benzo(a)pyrene. A second goal is to comprehend the configuration-dependent properties of these complexes in terms of basic chemical structure and conformation. Once agreement is established between the theoretically determined quantities and those measured in experiments, it may be possible to identify the structural characteristics of the carcinogens that underlie the unusual physical behavior of such modified polynucleotides. If is further hoped that these studies may offer some unifying principles for understanding the effects of carcinogen binding on nuclic acid conformation. With this knowledge as a conceptual basis, it then may be possible also to rationalize the changes in biological function that accompany the covalent binding of aromatic carcinogens to the nucleic acids.